期刊
IEEE TRANSACTIONS ON INDUSTRIAL INFORMATICS
卷 17, 期 5, 页码 3090-3098出版社
IEEE-INST ELECTRICAL ELECTRONICS ENGINEERS INC
DOI: 10.1109/TII.2020.3027746
关键词
Feature extraction; Nonlinear dynamical systems; Correlation; Microwave integrated circuits; Data models; Informatics; Machine learning; Maximal information coefficient (MIC); semisupervised learning; soft sensor; supervised weighted nonlinear dynamic system (WNDS); variational auto-encoder (VAE)
类别
资金
- National Key Research and Development Program of China [2018YFC0808600]
- National Natural Science Foundation of China (NSFC) [61833014, 61722310]
- Natural Science Foundation of Zhejiang Province [LR18F030001, TII-20-0400]
In this study, a soft sensor model based on deep learning is proposed, utilizing a supervised weighted nonlinear dynamic system model and the maximal information coefficient to predict key quality variables. By using a variational autoencoder to extract nonlinear dynamic features, the model can analyze correlations between variables and relationships between historical and present samples simultaneously.
Soft sensor plays an increasingly important role in modern industrial processes for estimating key quality variables which are hard to measure. With the development of deep learning technologies, soft sensors based on the deep learning methods have drawn great attention. Aiming to predict key quality variables, a supervised weighted nonlinear dynamic system (WNDS) model aided by the maximal information coefficient (MIC) is proposed in this article. The variational autoencoder is employed into the system for extracting nonlinear dynamic features. The supervised WNDS model can simultaneously analyze the correlations between variables and the relationships between historical samples and present samples. Furthermore, the proposed method is extended to a semisupervised form, in order to handle the imbalanced numbers between routinely recorded process data and limited labeled quality data. The prediction performance is validated by an industrial case.
作者
我是这篇论文的作者
点击您的名字以认领此论文并将其添加到您的个人资料中。
推荐
暂无数据